US 20060132811 A1
An improved method of encoding and compressing digital halftones that utilizes a “none-of-the-above” method for designating variable-length runs. A monochrome input bitmap is rearranged slightly to reduce the patterns possible in contained digital halftone cells. This revised monochrome bitmap is parsed into subfiles to optimize run-lengths. The parsed subfiles are combined into a single file whose alternating runs of 1's and 0's are converted into successive variable-length binary numbers. One of the permutations of an antecedent binary is designated “none-of-the-above” and its use triggers a subsequent variable-length binary. All other permutations within each variable-length binary may designate a specific contained run-length and such use triggers a return to the initial binary in the series. The above method is reversed to decode and uncompress the encoded file to reproduce the original revised monochrome bitmap for display by a computer monitor or printer.
1. A method for compressing and decompressing a digital halftone monochrome bitmap comprising the steps of:
(a) obtaining an input digital halftone monochrome bitmap of dimensions in pixels high and wide such that a plurality of equal-sized multi-pixel halftone cells are wholly contained,
(b) reducing the possible patterns of monochrome pixels within said halftone cells by shifting certain pixels either horizontally or vertically according to a master grid based on a checkerboard pattern,
(c) parsing said revised halftone cells into individual files, one each for each pixel position within the cell,
(d) recombining said files into a single file,
(e) encoding said single file's alternating runs of two monochrome colors with variable-length binary numbers within which all but one binary permutation may designate the length of the run and a return to start, and one binary permutation designates none-of-the-above contained run-lengths and to move on to a larger successive binary number.
(f) generating the increasing sizes of said successive run-length-containing variable-length binary numbers by a plurality of means, including Fibonacci numbers, byte-length-based numbers, or other means based on efficiency,
(g) generating a compressed file of fewer bits than the input bitmap by utilizing step e,
(h) reversing said method until said compressed file is uncompressed to replicate the said reduced pattern bitmap,
(i) displaying said bitmap on a computer printer or monitor.
1. Field of Invention
My invention relates to digital halftone compression and decompression, specifically to delineating variable-length runs using a “none-of-the-above” method.
2. Description of Prior Art
Digital halftone files have proven to be quite difficult to compress. Within the monochrome bitmap, each picture element, or pixel, is designated one of two “colors,” usually black or white. Due to the complexity of the image to be reproduced, the probability of any individual pixel being one color or the other is 50-50. This unpredictability affects the production of “runs,” large groups of pixels of the same “color.” Common compression methods based on run-length encoding prove inefficient, occasionally yielding “compressed” files even larger than the original input file.
Presently the digital halftone industry works around this problem by waiting until the last minute to create the monochrome bitmap. Complex image files utilizing a number of different file formats are compressed, stored or transmitted, and decompressed prior to being “rasterized” into a digital halftone file at the final output stage, the digital printer or monitor display.
Several objects and advantages of the present invention are:
(a.) to encode and compress a monochrome bitmap into a smaller binary file that my be efficiently stored or transmitted.
(b.) to decode and decompress this smaller binary file into a reproduction of the original monochrome bitmap that then may be displayed on a digital computer monitor or printer.
(c.) to store and transmit as a single file complex documents previously consisting of two or more different file formats.
(d.) to accomplish the above efficiently by means of a digital computer.
According to the present invention, a digital halftone efficiently is compressed for storage or transmission, and then efficiently decompressed for display on a digital computer monitor or printer.
A Monochrome Input Bitmap contains an input image's pixels displayed at a binary depth of one bit. Each pixel is described by either a 0 or a 1 to designate one of two “colors,” normally black or white. This bitmap is of certain dimensions in pixels high and wide, and a number of equal-sized multi-pixel halftone cells are contained wholly within.
A Revised Input Bitmap is derived by slightly rearranging the black and white pixels in each of the Monochrome Input Bitmap's contained two-pixel-by-two-pixel halftone cells to reduce the number of possible patterns. This pattern reduction maintains the ratio of black pixels to white pixels, or gray level, within the cell.
A Parsed Input Bitmap is derived by dividing the pixels within each contained two-pixel-by-two-pixel halftone cell into two or more separate bitmaps that maintain the spatial relationship of each pixel to similarly positioned pixels within the cell. These separate bitmaps then are combined into a global bitmap containing the same number of pixels as its predecessor and the same ratio of black pixels to white pixels, or global gray level.
A “none or the above” or NOTA Coder then acts on this Parsed Input Bitmap to yield a binary file that is smaller than that used to describe the preceding bitmaps.
Following retrieval of this file, a NOTA Decoder then acts to reverse the coding process and yield an exact duplicate of the Parsed Input Bitmap, now redesignated a Parsed Output Bitmap. This bitmap contains the same number of pixels and the same ratio of black pixels to white pixels as the Revised Input Bitmap but not in the same order.
A Revised Output Bitmap is derived by reversing the previous parsing method to yield a bitmap that is an exact duplicate of the Revised Input Bitmap, including the same number of global pixels and the same ratio of black pixels to white pixels within each two-by-two-pixel cell. This bitmap then is sent to a binary computer's Monitor or Printer Display.
From the description above, a number of advantages of my none-of-the-above digital halftone compression and decompression become evident.
(a) The actual size of the resultant encoded and compressed binary file is smaller than the size of the original monochrome bitmap file upon which it is based.
(b) Due to this reduction in size, the resultant file may be stored and transmitted more efficiently than the original file by using a binary computer.
(c) Unencoding and decompressing the resultant file provides a faithful reproduction of the revised input monochrome bitmap file.
(d) The none-of-the-above encoding and decoding method efficiently uses fewer resources of the binary computer system.
The subject of the operation of the current invention is an image expressed as a Monochrome Input Bitmap (
The smallest square digital halftone cell capable of depicting shades of gray measures two-pixels-by-two-pixels. While said cell may convey five shades, all pixels black plus the incremental addition of four white pixels, the number of possible patterns within the cell numbers sixteen. A method of the present invention is to reduce the number of allowed patterns within the two-pixel-by-two-pixel cells.
The four possible patterns for three pixels white, one pixel black (column 1, rows 2-5), are assigned to two allowed patterns (column 2, rows 2 & 3) by shifting the single black pixel horizontally as necessary to partially maintain the checkerboard pattern. The four possible patterns for one pixel white, three pixels black (column 3 rows 4-7), similarly are assigned to two allowed patterns (column 2, rows 5 & 6) by shifting the single white pixel horizontally as necessary to partially maintain the checkerboard pattern. (Note that the derived four patterns also may be created by moving the required pixels vertically and that the checkerboard pattern also may be the inverse.)
Once all the two-pixel-by-two-pixel cells similarly are processed, the Revised Input Bitmap (
This revised bitmap now is processed to yield several smaller bitmaps which are recombined into a Parsed Input Bitmap (
The four parsed files now are arranged sequentially in a file that may be visualized as a bitmap of dimensions half the number of pixels wide and twice the number of pixels high as the original input bitmap. Again, the global relationship of white to black pixels remains constant.
This Parsed Input Bitmap (
Observing the first run's color and length in
The smallest possible run is one pixel long. The smallest possible binomial that may depict such a run is one bit in length. Said one bit may depict two states, designated by one or zero. In this method, one of the states is the “none-of-the-above” designation which means that the current binomial does not contain the actual run-length and the next successive binomial must be examined. The second state is the actual run length which acts as a signal to start over again with the smallest possible binomial.
The size of the variable length binomials triggered by the NOTA state may be designated by any consecutive series of numbers. The preferred embodiment utilized in
Taking the first black run of four pixels in
Now, however, the run of four is contained in the third Fibonacci-length binomial of 2 bits. Thus the third and fourth positions of the variable length binomial are designated 10 (containing a1), which means “the second of the contained runs (3, 4, or 5 pixels), start over.” The run of four is thus designated by a four place binomial: 0010. A Compressed File equal in pixel width to the Revised Input Bitmap in
Taking the next white run of one pixel in
Taking the next black run of 31 pixels in
The run of 31 is greater than the runs of 3-5 pixels that may be contained in the next Fibonacci-length binomial of 2 bits. Thus the third and fourth positions of the variable length binomial are both designated 0 which means “none of the above contained runs (3-5 pixels), go on to the next binomial in the series.” The run of 31 also is greater than the runs of 6-12 pixels that may be contained in the next Fibonacci-length binomial of 3 bits. Thus the fifth, sixth and seventh positions of the variable length binomial are each designated 0 which means “none of the above contained runs (6-12 pixels), go on to the next binomial in the series.”
The run of 31, however, is contained in the next Fibonacci-length binomial of 5 bits. Thus the ninth through thirteenth positions of the variable length binomial are designated 10011 (containing a 1) which means “one of the contained runs (31 pixels), start over.” The run of 31 thus is designated by a 12 place binomial: 0000 0001 0011. In
The remaining portion of
Upon retrieval of said file, each encoding step is sequentially reversed. To recreate
As shown in
Also as shown in
This file then is displayed either on a personal computer screen or by a printer, designated as Screen or Printer Display (
Accordingly, the reader will see that the none-of-the-above method of this invention more efficiently compresses and decompresses a revised monochrome bitmap file. In addition, the method produces a compressed binary file that more efficiently may be stored and transmitted between digital computing devices. And, the method faithfully reproduces the revised input monochrome bitmap file for accurate display on a digital computer monitor or printer.
By using one permutation of a variable-length binary number to designate the parameters of a successive monochrome color run and any of the remaining per-mutations to designate the actual run and a return to the smallest variable-length binary, conservation of available resources of a binary computer relative to processing, storage and transmission is achieved. Such conservation relies on the method of this invention including the revision of the input monochrome bitmap and the parsing of the revised binary bitmap file to promote longer runs. The resulting compressed file results in a demonstrated saving over the size of the input file.
Although the description above contains many specificities, those should not be construed as limiting the scope of the invention but as merely providing illustration of some of the presently preferred embodiments of this invention.
Thus the scope of the invention should be determined by the appended claims and their legal equivalents, rather than by the examples given.